rev_gene()

Revise genes with rev_gene(). For scRNA-seq data, we suggest to revise the gene symbols with rev_gene(). geneinfo is the system data.frame containing the information of human and mouse from NCBI gene(updated in June. 19, 2022). To use your own geneinfo data.frame, please refer to demo_geneinfo() to build a new one, e.g., rat, zebrafish, Drosophila, C. elegans, etc.

Parameters of rev_gene() see below:

• data A data.frame or matrix or dgCMatrix
• data_type A character to define the type of data, select ‘count’ for the data matrix, ‘mir_info’ for the mir_info data.frame, ‘mir2tar’ for the mir2tar data.frame, ‘pathways’ for the pathways data.frame, ‘gene2go’ for the gene2go data.frame
• species Species of the data.’Human’, ‘Mouse’ or ‘Rat’
• geneinfo A data.frame of the system data containing gene symbols of ‘Human’, ‘Mouse’ and ‘Rat’ updated on June 19, 2022 for revising gene symbols

library(miRTalk)
#> 载入需要的程辑包：doParallel
#> Warning: 程辑包'doParallel'是用R版本4.1.3 来建造的
#> 载入需要的程辑包：foreach
#> 载入需要的程辑包：iterators
#> 载入需要的程辑包：parallel
#> Warning: replacing previous import 'Seurat::JS' by 'networkD3::JS' when loading
#> 'miRTalk'
load(paste0(system.file(package = "miRTalk"), "/extdata/example.rda"))

# demo_geneinfo
demo_geneinfo()
#>   symbol synonyms species
#> 1   A1BG      A1B   Human
#> 2   A1BG      ABG   Human
#> 3  A2MP1     A2MP   Human
#> 4   Aco1      Aco   Mouse
#> 5   Alb1      Alb     Rat

# demo_sc_data
demo_sc_data()
#> 6 x 6 sparse Matrix of class "dgCMatrix"
#>      cell1 cell2 cell3 cell4 cell5 cell6
#> A1BG     .    18     .     .     .    39
#> A2M      .     .    28     .     .     7
#> A2MP     .     .     5     .    26     .
#> NAT1     .     .    30     .     .    26
#> NAT2     .    29    13    42    25     .
#> NATP    14     .    15    19    38     .

# revise gene symbols
sc_data <- rev_gene(data = sc_data,data_type = "count",species = "Human",geneinfo = geneinfo)

create_miRTalk()

Create miRTalk object with create_miRTalk(). Users need to provide the raw count data and the cell type for create_miRTalk(). miRTalk will use NormalizeData() in Seurat to normalize data by default. If your data has been normalized by other methods, you can set the parameter if_normalize = FALSE

Parameters of create_miRTalk() see below: - sc_data mA data.frame or matrix or dgCMatrix containing raw counts of single-cell RNA-seq data - sc_celltype A character containing the cell type of the single-cell RNA-seq data - species A character meaning species of the single-cell transcriptomics data. “Human”, “Mouse”, “Rat” - if_normalize Normalize sc_data with Seurat LogNormalize()

obj <- create_miRTalk(sc_data = sc_data, sc_celltype = sc_celltype, species = "Human")

obj
#> An object of class miRTalk 
#> 0 EV-derived miRNA-target interactions

# If your data has been normalized by other methods, you can set the parameter `if_normalize = FALSE`

# obj <- create_miRTalk(sc_data = sc_data, sc_celltype = sc_celltype, species = "Human", if_normalize = FALSE)

find_miRNA()

Find expressed miRNAs among all cells with find_miRNA(). Users need to provided a EV-derived miRNA database for find_miRNA(). mir_info is the system data.frame containing information of EV-derived miRNA of “Human”, “Mouse”, and “Rat”. To use your own mir_info data.frame, please refer to demo_mir_info() to build a new one, e.g., zebrafish, Drosophila, C. elegans, etc.

Parameters of find_miRNA() see below: - object miRTalk object after create_miRTalk() - mir_info A data.frame of the system data containing information of EV-derived miRNA of “Human”, “Mouse”, and “Rat”

# demo_mir_info
demo_mir_info()
#>          miRNA    miRNA_mature    gene species
#> 1    hsa-miR-1    hsa-miR-1-5p  MIR1-1   Human
#> 2    hsa-miR-1    hsa-miR-1-5p  MIR1-2   Human
#> 3    hsa-miR-1    hsa-miR-1-3p  MIR1-1   Human
#> 4    hsa-miR-1    hsa-miR-1-3p  MIR1-2   Human
#> 5  mmu-miR-105     mmu-miR-105  Mir105   Mouse
#> 6 rno-miR-106b rno-miR-106b-5p Mir106b     Rat

# find highly expressed LR pairs
obj <- find_miRNA(object = obj, mir_info = mir_info)

find_hvtg()

Find highly variable target genes with DEGs and HVGs find_hvtg(). miRTalk uses the top 3000 HVGs and DEGs for each cell types with Seurat methods. The result is a character stored obj@data$var_genes. Users can use their own methods to find Find highly variable target genes and replace them

Parameters of find_hvtg() see below: - object miRTalk object after create_miRTalk() - pvalue Cutoff of p value. Default is 0.05 - log2fc log2 fold change for identifying the highly expressed genes in each cell type. Default is 0.5 - min_cell_num Min cell number for each cell type. Default is 10 - nfeatures Number of features to select as top variable features. Default is 3000

# find highly variable target genes
obj <- find_hvtg(object = obj)

# the result is a character
var_genes <- obj@data$var_genes
str(var_genes)
#>  chr [1:3377] "MGP" "LUM" "DCD" "DCN" "IGKV2D-28" "IGKV3-11" "SPINK1" ...

# if your have your own result, you can replace them with the following code

# object@data$var_genes <- your_own_var_genes

find_miRTalk()

Infer cell-cell communications mediated by EV-derived miRNAs from senders to receivers with find_miRTalk(). miRTalk uses the curated mir2tar database containing EV-derived miRNA-target interactions. To use your own mir2tar data.frame, please refer to demo_mir2tar() to build a new one, e.g., zebrafish, Drosophila, C. elegans, etc. The result is a character stored obj@cci. Users can use get_miRTalk_cci() to get simple results of miRNA-target interactions.

Parameters of find_miRTalk() see below: - object miRTalk object after find_miRNA() and find_hvtg() - mir2tar A data.frame of the system data containing relationship of miRNA and its target genes for “Human”, “Mouse”, “Rat” - min_cell_num Min cell number for each cell type and expressed miRNA. Default is 10 - pvalue Cutoff of p value. Default is 0.05 - resolution Correct to precursor or mature miRNAs. Use ‘precursor’ or ‘mature’. Default is ‘mature’ - min_percent Min percent of expressed cells for target genes of miRNA. Default is 0.05 - if_doParallel Use doParallel. Default is TRUE - use_n_cores umber of CPU cores to use. Default is 4

# demo_mir2tar
demo_mir2tar()
#>          miRNA    miRNA_mature target_gene species
#> 1    hsa-miR-1    hsa-miR-1-5p        BDNF   Human
#> 2    hsa-miR-1    hsa-miR-1-3p       RBM28   Human
#> 3  mmu-miR-105     mmu-miR-105        Abl2   Mouse
#> 4 rno-miR-106b rno-miR-106b-5p        Mcl1     Rat

# infer cell-cell communications mediated by EV-derived miRNAs from senders to receivers
obj <- find_miRTalk(object = obj, mir2tar = mir2tar)
obj
#> An object of class miRTalk 
#> 2185 EV-derived miRNA-target interactions

# the result is a data.frame
cci <- obj@cci
str(cci)
#> 'data.frame':    2185 obs. of  17 variables:
#>  $ celltype_sender           : chr  "Bcell" "Bcell" "Bcell" "Bcell" ...
#>  $ celltype_receiver         : chr  "Bcell" "Bcell" "Bcell" "Bcell" ...
#>  $ miRNA                     : chr  "hsa-miR-4426" "hsa-miR-29b-3p" "hsa-miR-29b-3p" "hsa-miR-29b-3p" ...
#>  $ miR_gene                  : chr  "MIR4426" "MIR29B1" "MIR29B1" "MIR29B1" ...
#>  $ percent_sender            : num  0.19 0.231 0.231 0.231 0.231 ...
#>  $ percent_receiver          : num  0.19 0.231 0.231 0.231 0.231 ...
#>  $ miRNA_activity            : num  0.429 0.481 0.481 0.481 0.481 ...
#>  $ target_gene               : chr  "PPIC" "CLDN1" "EREG" "TGFB3" ...
#>  $ target_gene_activity      : num  5.81e-03 3.45e-04 1.03e-04 3.43e-04 4.85e-05 ...
#>  $ target_gene_mean_exp      : num  0.02822 0.001677 0.000501 0.001667 0.000235 ...
#>  $ target_gene_mean_exp_other: num  0.187 0.1058 0.1496 0.1191 0.0943 ...
#>  $ target_gene_percent       : num  0.0964 0.0843 0.0964 0.0723 0.0843 ...
#>  $ target_gene_percent_other : num  0.403 0.208 0.211 0.303 0.164 ...
#>  $ pvalue                    : num  1.58e-08 2.63e-03 3.73e-03 1.85e-06 1.81e-02 ...
#>  $ sig                       : chr  "YES" "YES" "YES" "YES" ...
#>  $ prob                      : num  0.469 0.466 0.466 0.466 0.466 ...
#>  $ score                     : num  0.426 0.48 0.481 0.48 0.481 ...

# get simple results of miRNA-target interactions
obj_cci <- get_miRTalk_cci(obj)
str(obj_cci)
#> 'data.frame':    2083 obs. of  9 variables:
#>  $ celltype_sender     : chr  "Bcell" "Bcell" "Bcell" "Bcell" ...
#>  $ celltype_receiver   : chr  "Bcell" "Bcell" "Bcell" "Bcell" ...
#>  $ miRNA               : chr  "hsa-miR-4426" "hsa-miR-29b-3p" "hsa-miR-29b-3p" "hsa-miR-29b-3p" ...
#>  $ miR_gene            : chr  "MIR4426" "MIR29B1" "MIR29B1" "MIR29B1" ...
#>  $ miRNA_activity      : num  0.429 0.481 0.481 0.481 0.481 ...
#>  $ target_gene         : chr  "PPIC" "CLDN1" "EREG" "TGFB3" ...
#>  $ target_gene_activity: num  5.81e-03 3.45e-04 1.03e-04 3.43e-04 4.85e-05 ...
#>  $ prob                : num  0.469 0.466 0.466 0.466 0.466 ...
#>  $ score               : num  0.426 0.48 0.481 0.48 0.481 ...

plot_miRTalk_chord()

Parameters of plot_miRTalk_chord() see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot by order. Default is to plot all cell types
• celltype_color Colors for the cell types, whose length must be equal to celltype
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• edge_color Colors for the edges from the sender cell type, whose length must be equal to celltype
• edge_type Types for the edges from the sender cell type. Default is “big.arrow”. “ellipse” for ellipse, “triangle” for triangle, “curved” for curved. Details see circlize::chordDiagram()
• show_type which type of miRNAs to show, “number”, “activity”, or “score” for sum of inferred miRNAs number and activity, respectively, or “prob” for max probability. Default is “number”
• if_show_autocrine Whether to show autocrine. Default is FALSE
• text_size Size of text labels. Default is 1.5
• y_scale y_scale to adjust the text. Default is 0.1
• ... parameters pass to circlize::chordDiagram(), e.g., link.arr.width, link.arr.length, link.arr.col

plot_miRTalk_chord(object = obj)

plot_miRTalk_circle()

Parameters of plot_miRTalk_circle() see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot. Default is to plot all cell types
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• celltype_color Colors for the cell types, whose length must be equal to celltype
• edge_color Colors for the edges from the sender cell type, whose length must be equal to celltype
• edge_type Types for the edges. “fan” by default, “link”, “hive”
• show_type which type of miRNAs to show, “number”, “activity”, or “score” for sum of inferred miRNAs number and activity, respectively, or “prob” for max probability. Default is “number”
• if_show_autocrine Whether to show autocrine. Default is FALSE
• edge_alpha Transparency of edge. Default is 0.5
• node_size Size of node. Default is 10
• text_size Size of text. Default is 5

plot_miRTalk_circle(object = obj)

plot_miRTalk_circle_simple()

# as sender
plot_miRTalk_circle_simple(object = obj, celltype = "Tumor", celltype_dir = "sender")

# as receiver
plot_miRTalk_circle_simple(object = obj, celltype = "Tumor", celltype_dir = "reciver")

plot_miRTalk_heatmap()

Parameters of plot_miRTalk_heatmap() see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot. Default is to plot all cell types
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• show_type which type of miRNAs to show, “number”, “activity”, or “score” for sum of inferred miRNAs number and activity, respectively, or “prob” for max probability. Default is “number”
• text_size Size of text labels. Default is 10
• viridis_option option in viridis::scale_color_viridis, can be “A”, “B”, “C”, “D”, “E”, “F”, “G”, “H”. Default is “D”
• ... parameters pass to heatmaply::heatmaply, e.g., grid_color, grid_width

plot_miRTalk_heatmap(object = obj)

plot_miRTalk_sankey()

Parameters of plot_miRTalk_sankey() see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot by order. Default is to plot all cell types
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• celltype_color Colors for the cell types, whose length must be equal to celltype
• edge_color Colors for the edges from the sender cell type, whose length must be equal to celltype, Or use “NO” to cancel it
• show_type which type of miRNAs to show, “number”, “activity”, or “score” for sum of inferred miRNAs number and activity, respectively, or “prob” for max probability. Default is “number”
• if_show_autocrine Whether to show autocrine. Default is FALSE
• edge_alpha Transparency of edge. Default is 0.5
• node_size Size of node. Default is 40
• text_size Size of text. Default is 15
• node_pad Size of node padding. Numeric essentially influences the width height. Default is 20
• ... parameters pass to networkD3::sankeyNetwork

plot_miRTalk_sankey(object = obj)

plot_miR_heatmap()

Parameters of plot_miR_heatmap()see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot. Default is to plot all cell types
• miRNA which miRNAs to plot. Default is to plot all inferred miRNAs
• text_size Size of text labels. Default is 10
• if_horizontal Whether to plot with the horizontal direction. Default is TRUE
• viridis_option option in viridis::scale_color_viridis, can be “A”, “B”, “C”, “D”, “E”, “F”, “G”, “H”. Default is “D”.
• ... parameters pass to heatmaply::heatmaply, e.g., grid_color, grid_width

plot_miR_heatmap(object = obj)

plot_miR_bubble()

Parameters of plot_miR_bubble() see below:

• object miRTalk object after find_miRTalk()
• celltype which cell types to plot. Default is to plot all cell types
• miRNA which miRNAs to plot. Default is to plot all inferred miRNAs
• if_show_autocrine Whether to show autocrine. Default is FALSE
• if_horizontal Whether to plot with the horizontal direction. Default is TRUE
• viridis_option option in viridis::scale_color_viridis, can be “A”, “B”, “C”, “D”, “E”, “F”, “G”, “H”. Default is “D”.

plot_miR_bubble(object = obj)

plot_miR2tar_chord()

Parameters of plot_miR2tar_chord() see below:

• object miRTalk object after find_miRTalk()
• celltype_sender Name of celltype_sender. One or more cell types
• celltype_receiver Name of celltype_receiver. One or more cell types
• celltype_color Colors for the celltype_sender nodes and celltype_receiver nodes, or use “NO” to make it simple
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• edge_color Colors for the edges from the sender cell type
• edge_type Types for the edges from the sender cell type. Default is “circle”, “big.arrow” for big arrow, “triangle” for triangle, “ellipse” for ellipse, “curved” for curved. Details see circlize::chordDiagram
• show_type which type of miRNAs to show, “prob” or “score” for inferred miRNAs-target interactions. Default is “prob”
• text_size Size of text labels. Default is 0.5
• y_scale y_scale to adjust the text. Default is 1
• ... parameters pass to circlize::chordDiagram, e.g., link.arr.width, link.arr.length, link.arr.col

plot_miR2tar_chord(obj, celltype_sender = "Tumor", celltype_receiver = "Stromal")

plot_miR2tar_circle()

Parameters of plot_miR2tar_circle() see below:

• object miRTalk object after find_miRTalk()
• celltype_sender Name of celltype_sender. One or more cell types
• celltype_receiver Name of celltype_receiver. One or more cell types
• celltype_color Colors for the celltype_sender nodes and celltype_receiver nodes, or use “NO” to make it simple
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• edge_color Colors for the edges from the sender cell type
• text_size Size of text labels. Default is 3
• edge_width y_scale to adjust the text. Default is 0.5

plot_miR2tar_circle(obj, celltype_sender = "Tumor", celltype_receiver = "Stromal")

plot_miR2tar_circle(obj, celltype_sender = "Tumor", celltype_receiver = "Stromal", celltype_color = "NO")

plot_miR2tar_heatmap()

Parameters of plot_miR2tar_heatmap() see below:

• object miRTalk object after find_miRTalk()
• celltype_sender Name of celltype_sender. One or more cell types
• celltype_receiver Name of celltype_receiver. One or more cell types
• miRNA which miRNAs to use. Default is to plot all inferred miRNAs
• show_type which type of miRNAs to show, “prob” or “score” for inferred miRNAs-target interactions. Default is “prob”
• text_size Size of text labels. Default is 3
• if_horizontal Whether to plot with the horizontal direction. Default is TRUE
• viridis_option option in viridis::scale_color_viridis, can be “A”, “B”, “C”, “D”, “E”, “F”, “G”, “H”. Default is “D”.
• ... parameters pass to heatmaply::heatmaply, e.g., grid_color

plot_miR2tar_heatmap(obj, celltype_sender = "Tumor", celltype_receiver = "Stromal", grid_color = "black")

To plot the sparse expression of miRNA genes in scRNA-seq, you can use the Nebulosa::plot_density()

Nebulosa::plot_density(object = obj_seurat, features = "MIR24-2")

sessionInfo()
#> R version 4.1.1 (2021-08-10)
#> Platform: x86_64-w64-mingw32/x64 (64-bit)
#> Running under: Windows 10 x64 (build 19045)
#> 
#> Matrix products: default
#> 
#> locale:
#> [1] LC_COLLATE=Chinese (Simplified)_China.936 
#> [2] LC_CTYPE=Chinese (Simplified)_China.936   
#> [3] LC_MONETARY=Chinese (Simplified)_China.936
#> [4] LC_NUMERIC=C                              
#> [5] LC_TIME=Chinese (Simplified)_China.936    
#> 
#> attached base packages:
#> [1] parallel  stats     graphics  grDevices utils     datasets  methods  
#> [8] base     
#> 
#> other attached packages:
#> [1] miRTalk_1.0       doParallel_1.0.17 iterators_1.0.13  foreach_1.5.1    
#> 
#> loaded via a namespace (and not attached):
#>   [1] circlize_0.4.13       plyr_1.8.6            igraph_1.2.7         
#>   [4] lazyeval_0.2.2        sp_1.4-5              splines_4.1.1        
#>   [7] crosstalk_1.2.0       listenv_0.8.0         scattermore_0.7      
#>  [10] ggplot2_3.3.6         digest_0.6.28         ca_0.71.1            
#>  [13] htmltools_0.5.2       viridis_0.6.2         fansi_0.5.0          
#>  [16] magrittr_2.0.1        tensor_1.5            cluster_2.1.2        
#>  [19] ROCR_1.0-11           limma_3.50.0          globals_0.14.0       
#>  [22] graphlayouts_0.7.1    matrixStats_0.61.0    spatstat.sparse_3.0-0
#>  [25] prettyunits_1.1.1     rmdformats_1.0.3      colorspace_2.0-2     
#>  [28] ggrepel_0.9.1         xfun_0.30             dplyr_1.0.7          
#>  [31] crayon_1.4.2          jsonlite_1.7.2        progressr_0.9.0      
#>  [34] spatstat.data_3.0-0   survival_3.2-11       zoo_1.8-9            
#>  [37] glue_1.4.2            polyclip_1.10-0       registry_0.5-1       
#>  [40] gtable_0.3.0          webshot_0.5.4         leiden_0.3.9         
#>  [43] future.apply_1.8.1    shape_1.4.6           abind_1.4-5          
#>  [46] scales_1.1.1          pheatmap_1.0.12       DBI_1.1.1            
#>  [49] miniUI_0.1.1.1        Rcpp_1.0.7            progress_1.2.2       
#>  [52] viridisLite_0.4.0     xtable_1.8-4          reticulate_1.22      
#>  [55] spatstat.core_2.3-0   datawizard_0.6.2      htmlwidgets_1.5.4    
#>  [58] httr_1.4.2            RColorBrewer_1.1-2    ellipsis_0.3.2       
#>  [61] Seurat_4.1.1          ica_1.0-2             pkgconfig_2.0.3      
#>  [64] farver_2.1.0          sass_0.4.0            uwot_0.1.10          
#>  [67] deldir_1.0-6          utf8_1.2.2            labeling_0.4.2       
#>  [70] tidyselect_1.1.1      rlang_0.4.12          reshape2_1.4.4       
#>  [73] later_1.3.0           munsell_0.5.0         tools_4.1.1          
#>  [76] generics_0.1.1        ggridges_0.5.3        evaluate_0.14        
#>  [79] stringr_1.4.0         fastmap_1.1.0         heatmaply_1.4.0      
#>  [82] yaml_2.2.1            goftest_1.2-3         knitr_1.36           
#>  [85] fitdistrplus_1.1-6    tidygraph_1.2.0       purrr_0.3.4          
#>  [88] RANN_2.6.1            dendextend_1.16.0     ggraph_2.0.5         
#>  [91] pbapply_1.5-0         future_1.23.0         nlme_3.1-152         
#>  [94] mime_0.12             correlation_0.8.3     compiler_4.1.1       
#>  [97] plotly_4.10.0         png_0.1-7             spatstat.utils_3.0-1 
#> [100] tibble_3.1.5          tweenr_1.0.2          bslib_0.3.1          
#> [103] stringi_1.7.5         highr_0.9             rgeos_0.5-9          
#> [106] lattice_0.20-44       Matrix_1.3-4          vctrs_0.3.8          
#> [109] pillar_1.6.4          lifecycle_1.0.1       networkD3_0.4        
#> [112] spatstat.geom_3.0-3   lmtest_0.9-38         jquerylib_0.1.4      
#> [115] GlobalOptions_0.1.2   RcppAnnoy_0.0.19      insight_0.18.5       
#> [118] data.table_1.14.2     cowplot_1.1.1         irlba_2.3.3          
#> [121] seriation_1.4.0       httpuv_1.6.3          patchwork_1.1.1      
#> [124] R6_2.5.1              bookdown_0.25         promises_1.2.0.1     
#> [127] TSP_1.2-1             KernSmooth_2.23-20    gridExtra_2.3        
#> [130] parallelly_1.28.1     codetools_0.2-18      MASS_7.3-54          
#> [133] assertthat_0.2.1      SeuratObject_4.1.0    sctransform_0.3.3    
#> [136] bayestestR_0.13.0     hms_1.1.1             mgcv_1.8-36          
#> [139] grid_4.1.1            rpart_4.1-15          tidyr_1.1.4          
#> [142] rmarkdown_2.13        Rtsne_0.15            ggforce_0.3.3        
#> [145] shiny_1.7.1